Table 1.

Mean (standard deviation) of the outcome variance, bias and percent bias of the estimated outcome variance across 200 simulations for different window size, total sample size and outcome prevalence. Bias is defined as the true minus estimated variance. Percent bias is defined as the true minus estimated variance divided by the true variance.

	n = 200			n = 400			n = 800
Window size	p = 0.1	p = 0.3	p = 0.5	p = 0.1	p = 0.3	p = 0.5	p = 0.1	p = 0.3	p = 0.5
True value
	0.0850 (0.0155)	0.1764 (0.0137)	0.2052 (0.0119)	0.0861 (0.0107)	0.1767 (0.0103)	0.2058 (0.0083)	0.0851 (0.0074)	0.1780 (0.0067)	0.2055 (0.0059)
Bias
10	0.0066 (0.0036)	0.0141 (0.0050)	0.0161 (0.0061)	0.0068 (0.0024)	0.0140 (0.0043)	0.0166 (0.0046)	0.0066 (0.0019)	0.0143 (0.0029)	0.0164 (0.0035)
20	0.0029 (0.0032)	0.0058 (0.0044)	0.0065 (0.0052)	0.0030 (0.0022)	0.0059 (0.0038)	0.0073 (0.0039)	0.0029 (0.0018)	0.0064 (0.0026)	0.0073 (0.0034)
30	0.0014 (0.0031)	0.0021 (0.0044)	0.0022 (0.0052)	0.0017 (0.0021)	0.0029 (0.0037)	0.0037 (0.0036)	0.0015 (0.0018)	0.0035 (0.0025)	0.0041 (0.0033)
40	0.0006 (0.0031)	−0.0002 (0.0043)	−0.0006 (0.0051)	0.0009 (0.0020)	0.0012 (0.0036)	0.0018 (0.0035)	0.0008 (0.0018)	0.0020 (0.0025)	0.0024 (0.0032)
Percent bias
10	0.0792 (0.0442)	0.0799 (0.0272)	0.0785 (0.0296)	0.0792 (0.0276)	0.0789 (0.0237)	0.0808 (0.0219)	0.0774 (0.0217)	0.0803 (0.0157)	0.0798 (0.0166)
20	0.0350 (0.0411)	0.0325 (0.0249)	0.0314 (0.0258)	0.0351 (0.0256)	0.0334 (0.0216)	0.0353 (0.0188)	0.0339 (0.0209)	0.0358 (0.0144)	0.0356 (0.0162)
30	0.0182 (0.0417)	0.0117 (0.0248)	0.0109 (0.0256)	0.0190 (0.0246)	0.0163 (0.0209)	0.0181 (0.0177)	0.0182 (0.0208)	0.0198 (0.0142)	0.0198 (0.0161)
40	0.0088 (0.0423)	−0.0016 (0.0247)	−0.0033 (0.0254)	0.0102 (0.0238)	0.0064 (0.0208)	0.0085 (0.0172)	0.0099 (0.0207)	0.0112 (0.0143)	0.0114 (0.0159)